{"paper":{"title":"You Can't Fool Us: Understanding the Resilience of LLM-driven Agent Communities to Misinformation","license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","headline":"Higher open-minded thinking in simulated communities reduces misinformation uptake and speeds recovery, while polarization leaves more lingering support.","cross_cats":[],"primary_cat":"cs.CY","authors_text":"Chichen Lin, Han Xiao, Kangbo Hu, Weijian Fan, Yijie Jin, Yongbin Wang, Zhanzhan Zhao, Zhihui Ying","submitted_at":"2026-05-17T09:45:33Z","abstract_excerpt":"Misinformation resilience is a dynamic community process: communities differ not only in whether they initially trust false claims, but also in how they recover through interaction, questioning, correction, and support withdrawal. We study this process with an LLM-based agent simulation that constructs synthetic communities along two theoretically motivated dimensions: Actively Open-minded Thinking (AOT), which captures evidence-seeking and willingness to revise beliefs, and Political Ideology (PI), which captures identity-based interpretation of contested claims. These two traits allow us to "},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"Across systematically varied AOT-PI communities, we find that higher AOT improves both resistance to misinformation uptake and recovery after trust peaks. PI shapes the recovery pathway: ideologically moderate communities recover more reliably, while polarized communities retain more residual support.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"LLM-based agents can faithfully simulate human psychological processes and social interactions when assigned traits such as Actively Open-minded Thinking and Political Ideology in response to misinformation shocks.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"LLM agent simulations show higher actively open-minded thinking boosts resistance to and recovery from misinformation while ideological moderation supports more reliable correction than polarization.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"Higher open-minded thinking in simulated communities reduces misinformation uptake and speeds recovery, while polarization leaves more lingering support.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"f8dbb4e69e6f5c81c290b16c45314fd379b78e64e996a5d199624c84e0daaa10"},"source":{"id":"2605.17353","kind":"arxiv","version":1},"verdict":{"id":"a971bd3f-168b-4a73-863a-fc5cd614b125","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-19T23:14:09.041291Z","strongest_claim":"Across systematically varied AOT-PI communities, we find that higher AOT improves both resistance to misinformation uptake and recovery after trust peaks. PI shapes the recovery pathway: ideologically moderate communities recover more reliably, while polarized communities retain more residual support.","one_line_summary":"LLM agent simulations show higher actively open-minded thinking boosts resistance to and recovery from misinformation while ideological moderation supports more reliable correction than polarization.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"LLM-based agents can faithfully simulate human psychological processes and social interactions when assigned traits such as Actively Open-minded Thinking and Political Ideology in response to misinformation shocks.","pith_extraction_headline":"Higher open-minded thinking in simulated communities reduces misinformation uptake and speeds recovery, while polarization leaves more lingering support."},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2605.17353/integrity.json","findings":[],"available":true,"detectors_run":[{"name":"doi_title_agreement","ran_at":"2026-05-19T23:31:20.086397Z","status":"completed","version":"1.0.0","findings_count":0},{"name":"doi_compliance","ran_at":"2026-05-19T23:21:01.504537Z","status":"completed","version":"1.0.0","findings_count":0},{"name":"claim_evidence","ran_at":"2026-05-19T21:41:57.792359Z","status":"completed","version":"1.0.0","findings_count":0},{"name":"ai_meta_artifact","ran_at":"2026-05-19T21:33:23.724964Z","status":"skipped","version":"1.0.0","findings_count":0}],"snapshot_sha256":"6d028b87e3e83497fcb071d070ea3bdee0a981234f15c7e652c4abd8e80aa856"},"references":{"count":13,"sample":[{"doi":"","year":2026,"title":"Borah, A.; Mihalcea, R.; and Perez-Rosas, V","work_id":"59985531-3dd6-4f1c-b5d6-b23d94466c2b","ref_index":1,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2021,"title":"LLM Agents Grounded in Self-Reports Enable General-Purpose Simulation of Individuals","work_id":"b8eecc06-3e78-4eba-9e1e-7c3bbebeaa4d","ref_index":2,"cited_arxiv_id":"2411.10109","is_internal_anchor":true},{"doi":"","year":null,"title":"Roozenbeek, J.; Freeman, A","work_id":"7b74f14c-4573-485f-8025-b471981172ab","ref_index":3,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2019,"title":"Beyond the Crowd: LLM-Augmented Community Notes for Governing Health Misinformation","work_id":"70f12fb5-4d5d-4ad9-b866-d9bb69996437","ref_index":4,"cited_arxiv_id":"2510.11423","is_internal_anchor":true},{"doi":"","year":null,"title":"For most authors... (a) Would answering this research question advance sci- ence without violating social contracts, such as violat- ing privacy norms, perpetuating unfair profiling, exac- erbating th","work_id":"68f63f59-61a0-4bf8-9275-2163dc024241","ref_index":5,"cited_arxiv_id":"","is_internal_anchor":false}],"resolved_work":13,"snapshot_sha256":"aed938ed57432cf1547ff36a00fa822bb3799b31738a5fc0e9483c60af2d534b","internal_anchors":2},"formal_canon":{"evidence_count":2,"snapshot_sha256":"6bc0354e02da9d28226ce01b1ebfe436efefb1c8a1e8cee75e2dbbda1119b05d"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"}